Final Project¶

Data Collection¶

rm(list=ls())

options(warn=-1)
library(tidyverse)

-- Attaching packages --------------------------------------- tidyverse 1.3.0 --

v ggplot2 3.3.2     v purrr   0.3.4
v tibble  3.0.3     v dplyr   1.0.2
v tidyr   1.1.2     v stringr 1.4.0
v readr   1.3.1     v forcats 0.5.0

-- Conflicts ------------------------------------------ tidyverse_conflicts() --
x dplyr::filter() masks stats::filter()
x dplyr::lag()    masks stats::lag()

Now, we will read all the necessory CSV files and save the data which we need to use.

sales = read_csv("sales_train.csv") 
head(sales)

Parsed with column specification:
cols(
  date = col_character(),
  date_block_num = col_double(),
  shop_id = col_double(),
  item_id = col_double(),
  item_price = col_double(),
  item_cnt_day = col_double()
)

items = read_csv("items.csv")
head(items)

Parsed with column specification:
cols(
  item_name = col_character(),
  item_id = col_double(),
  item_category_id = col_double()
)

shops = read_csv("shops.csv")
head(shops)

Parsed with column specification:
cols(
  shop_name = col_character(),
  shop_id = col_double()
)

item_categories = read_csv("item_categories.csv" )
head(item_categories)

Parsed with column specification:
cols(
  item_category_name = col_character(),
  item_category_id = col_double()
)

# Joining the items, shops and item_categories data to the sales data

merged_sales = sales %>% 
inner_join(items, by = "item_id") %>%
inner_join(shops, by = "shop_id") %>%
inner_join(item_categories, by = "item_category_id")
head(merged_sales)

# Seprate dates into month, day, year

library(lubridate) 
merged_sales$date = dmy(merged_sales$date)

merged_sales$year = year(merged_sales$date)
merged_sales$month = month(merged_sales$date)
merged_sales$day = day(merged_sales$date)
merged_sales$weekday = weekdays(merged_sales$date)

Attaching package: 'lubridate'


The following objects are masked from 'package:base':

    date, intersect, setdiff, union

# Converting the columns into factors

merged_sales$year = as.factor(merged_sales$year)
merged_sales$month = as.factor(merged_sales$month)
merged_sales$weekday = as.factor(merged_sales$weekday)

merged_sales$shop_id = as.factor(merged_sales$shop_id)
merged_sales$item_id = as.factor(merged_sales$item_id)
merged_sales$item_category_id = as.factor(merged_sales$item_category_id)

str(merged_sales)

tibble [2,935,849 x 14] (S3: spec_tbl_df/tbl_df/tbl/data.frame)
 $ date              : Date[1:2935849], format: "2013-01-02" "2013-01-03" ...
 $ date_block_num    : num [1:2935849] 0 0 0 0 0 0 0 0 0 0 ...
 $ shop_id           : Factor w/ 60 levels "0","1","2","3",..: 60 26 26 26 26 26 26 26 26 26 ...
 $ item_id           : Factor w/ 21807 levels "0","1","2","3",..: 21792 2496 2496 2498 2499 2508 2509 2515 2515 2516 ...
 $ item_price        : num [1:2935849] 999 899 899 1709 1099 ...
 $ item_cnt_day      : num [1:2935849] 1 1 -1 1 1 1 1 1 1 3 ...
 $ item_name         : chr [1:2935849] "<U+042F><U+0412><U+041B><U+0415><U+041D><U+0418><U+0415> 2012 (BD)" "DEEP PURPLE  The House Of Blue Light  LP" "DEEP PURPLE  The House Of Blue Light  LP" "DEEP PURPLE  Who Do You Think We Are  LP" ...
 $ item_category_id  : Factor w/ 84 levels "0","1","2","3",..: 38 59 59 59 57 60 57 56 56 56 ...
 $ shop_name         : chr [1:2935849] "<U+042F><U+0440><U+043E><U+0441><U+043B><U+0430><U+0432><U+043B><U+044C> <U+0422><U+0426> \"<U+0410><U+043B><U+"| __truncated__ "<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0420><U+041A> \"<U+0410><U+0442><U+0440><U+0438><U+0443><U+043C>\"" "<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0420><U+041A> \"<U+0410><U+0442><U+0440><U+0438><U+0443><U+043C>\"" "<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0420><U+041A> \"<U+0410><U+0442><U+0440><U+0438><U+0443><U+043C>\"" ...
 $ item_category_name: chr [1:2935849] "<U+041A><U+0438><U+043D><U+043E> - Blu-Ray" "<U+041C><U+0443><U+0437><U+044B><U+043A><U+0430> - <U+0412><U+0438><U+043D><U+0438><U+043B>" "<U+041C><U+0443><U+0437><U+044B><U+043A><U+0430> - <U+0412><U+0438><U+043D><U+0438><U+043B>" "<U+041C><U+0443><U+0437><U+044B><U+043A><U+0430> - <U+0412><U+0438><U+043D><U+0438><U+043B>" ...
 $ year              : Factor w/ 3 levels "2013","2014",..: 1 1 1 1 1 1 1 1 1 1 ...
 $ month             : Factor w/ 12 levels "1","2","3","4",..: 1 1 1 1 1 1 1 1 1 1 ...
 $ day               : int [1:2935849] 2 3 5 6 15 10 2 4 11 3 ...
 $ weekday           : Factor w/ 7 levels "Friday","Monday",..: 7 5 3 4 6 5 7 1 1 5 ...
 - attr(*, "spec")=
  .. cols(
  ..   date = col_character(),
  ..   date_block_num = col_double(),
  ..   shop_id = col_double(),
  ..   item_id = col_double(),
  ..   item_price = col_double(),
  ..   item_cnt_day = col_double()
  .. )

# Check for missing data 

any(is.na(merged_sales))

This is necessory just to make sure we don't have any NAs which can cause problems in the future.

Feature Choice¶

We will visualise the data first in order the find out which features are important.¶

#Total items by month

aggregate(item_cnt_day~month, merged_sales, sum)

# This will give number of distinct shops 

no_of_shops = merged_sales %>% 
select(shop_id) %>% 
distinct() %>% 
summarise(shops=n())

no_of_shops

# This will give number of times a shop apperars in the data set.

shop_frequency = merged_sales %>% 
group_by(shop_id,shop_name) %>% 
summarise(number=n())

shop_frequency

`summarise()` regrouping output by 'shop_id' (override with `.groups` argument)

# The will give the most popular shop (i.e. by sales volume)

popular_shop = merged_sales %>% 
group_by(shop_id,shop_name) %>%
summarise(total = sum(item_cnt_day)) %>%
arrange(desc(total))

popular_shop

`summarise()` regrouping output by 'shop_id' (override with `.groups` argument)

library(plotly)

Attaching package: 'plotly'


The following object is masked from 'package:ggplot2':

    last_plot


The following object is masked from 'package:stats':

    filter


The following object is masked from 'package:graphics':

    layout

# Plot for the shop id vs total quantity sold

plot_ly(popular_shop, x=~total, y=~reorder(as.factor(shop_id) , total), 
        type="bar", color=~shop_id, hoverinfo='text', text=~paste(shop_name,':',total), orientation='h') %>%
layout(title="Most Popular Shop by Total Quantity Sold.", xaxis=list(title="Total Quantity Sold"), 
       yaxis=list(title="Shop ID"),legend=list(title=list(text='<b> Shop ID </b>')))

# Unique number of items 

unique_items = merged_sales %>% 
distinct(item_id) %>%
summarise(items=n())

unique_items

item_freq = merged_sales %>% 
group_by(item_id,item_name) %>%
summarise(freq =n()) %>%
arrange(desc(freq))

head(item_freq)

`summarise()` regrouping output by 'item_id' (override with `.groups` argument)

# Most sold item for each shop

popular_item = merged_sales %>% 
group_by(shop_id, item_id) %>%
summarise(quantity_sold = sum(item_cnt_day)) %>%
filter(quantity_sold == max(quantity_sold))


popular_item

`summarise()` regrouping output by 'shop_id' (override with `.groups` argument)

# Plot for the most quantity sold for each shop

plot_ly(popular_item,y=~shop_id, x=~quantity_sold,
        type="bar", color=~as.factor(item_id), orientation='h', hoverinfo='text', text=~paste('Shop',shop_id,':','Item',item_id)) %>%
    layout(title="Most Popular Item per Shop", yaxis=list(title="Shop"), xaxis=list(title="Quantity Sold"), legend=list(title=list(text='<b> Item ID </b>')))

# Repetition of categories in the data set 

item_cat_freq = merged_sales %>% 
    group_by(item_category_id,item_category_name) %>%
    summarise(freq =n()) %>%
    arrange(desc(freq))

head(item_cat_freq)

`summarise()` regrouping output by 'item_category_id' (override with `.groups` argument)

# Find number of unique categories

item_cat_unique = merged_sales %>%
                select(item_category_id) %>% 
                distinct() %>% 
                count()

item_cat_unique

# Find number of categories per shop

categories_shop = merged_sales %>% 
                group_by(shop_id) %>%
                distinct(item_category_id) %>% 
                summarise(no_distinct_categories = n()) 

categories_shop

`summarise()` ungrouping output (override with `.groups` argument)

# Plot for shop vs number of categories for each shop

plot_ly(categories_shop, x=~no_distinct_categories, y=~reorder(as.factor(shop_id), no_distinct_categories),
       type='bar', orientation='h', color=~as.factor(shop_id), hoverinfo='text', text=~paste('Shop',shop_id,':',no_distinct_categories,'Categories')) %>%
    layout(title= "Number of categories per shop", xaxis=list(title= "Number of distinct categories"), yaxis=list(title= "Shops"), showlegend=FALSE)

#  Find top 10 selling products

top10_prod = merged_sales %>% 
            group_by(item_id,item_name) %>%
            summarise(qty_vol = sum(item_cnt_day)) %>%
            arrange(desc(qty_vol)) %>%
            ungroup() %>%
            top_n(10)

top10_prod

`summarise()` regrouping output by 'item_id' (override with `.groups` argument)

Selecting by qty_vol

# Plot for the quantity vs item for the top 10 selling items

plot_ly(top10_prod, x=~item_name, y=~qty_vol, type='bar', color=~as.factor(item_id), hoverinfo='text', text=~paste(item_name,':',qty_vol)) %>%
    layout(title="Top 10 Selling Products", xaxis=list(title="Item ID"), yaxis=list(title="Quantity Sold"), showlegend=FALSE)

# Find top 10 selling categories

top10_cat = merged_sales %>% 
            group_by(item_category_id,item_category_name) %>%
            summarise(qty_vol = sum(item_cnt_day)) %>%
            arrange(desc(qty_vol)) %>%
            ungroup() %>%
            top_n(10)

top10_cat

`summarise()` regrouping output by 'item_category_id' (override with `.groups` argument)

Selecting by qty_vol

# Plot for quantity vs categories for the top 10 selling categories

plot_ly(top10_cat, x=~item_category_name, y=~qty_vol, color=~as.factor(item_category_id), type='bar',  hoverinfo='text', text=~paste(item_category_name,':',qty_vol)) %>%
    layout(title="Top 10 Selling Categories", xaxis=list(title="Category"), yaxis=list(title="Quantity Sold"), showlegend=FALSE)

#  Monthly sales for each year

year_month_sales = merged_sales %>% 
                   group_by(year, month) %>%
                   summarise(sales_vol = sum(item_cnt_day * item_price)) %>%
                   arrange(year)

year_month_sales

`summarise()` regrouping output by 'year' (override with `.groups` argument)

# Plot for sales by volume vs year with montly bifurcation

plot_ly(year_month_sales, x=~year, y=~sales_vol, color=~month, type='bar', hoverinfo='text', text=~paste(month.name[month],':',round(sales_vol))) %>%
        layout(title="Sales year-month Wise", xaxis=list(title="Year"), yaxis=list(title="Sales Volume"), barmode='stack', legend=list(title=list(text='<b>Month</b>')))

#  Total yearly sales

year_sales = merged_sales %>% 
            group_by(year) %>%
            summarise(sales_vol_year = sum(item_cnt_day * item_price)) 

year_sales

`summarise()` ungrouping output (override with `.groups` argument)

# Plot for total sales vs year

plot_ly(year_sales, x=~year, y=~sales_vol_year, type='bar', color=~year, hoverinfo='text', text=~paste(year,':',round(sales_vol_year))) %>%
    layout(title="Total sales per Year", xaxis=list(title="Year"), yaxis=list(title="Total Sales"), legend=list(title=list(text='<b>Year</b>')))

# Find the best selling product by quantity

prod_popularity_years = merged_sales %>% 
                        group_by(year, item_id) %>%
                        summarise(qty_vol = sum(item_cnt_day)) %>%
                        filter(qty_vol == max(qty_vol))

prod_popularity_years

`summarise()` regrouping output by 'year' (override with `.groups` argument)

# Plot for the quantity of the most sold product vs year

plot_ly(prod_popularity_years, x=~year, y=~qty_vol, type='bar', hoverinfo='text', text=~paste(year,', Item',item_id,':',qty_vol)) %>%
    layout(title="Most Popular Product per year", xaxis=~list(title="Year"), yaxis=~list(title="Quantity Sold"))

# Find the shop with highest sales volume

sales_volume = merged_sales %>% 
              group_by(shop_id,shop_name) %>%
              summarise(sales_vol = sum(item_cnt_day * item_price)) %>%
              arrange(desc(sales_vol))

sales_volume

`summarise()` regrouping output by 'shop_id' (override with `.groups` argument)

# Plot for shop vs sales volume

plot_ly(sales_volume, y=~reorder(as.factor(shop_id) , sales_vol), x=~sales_vol, color=~as.factor(shop_id), type='bar', hoverinfo='text', text=~paste('Shop',shop_id,':',round(sales_vol)), orientation='h') %>%
    layout(title="Sales Volume", yaxis=list(title="Shop"), xaxis=list(title="Sales Volume(Amount)"), legend=list(title=list(text='<b>Shop ID</b>')))

After all the visualisation we can determine that 'month' and 'item_category_id' are important features and should be considered to be used for training our model.¶

# We will now convert our dataframe into a data table which makes it easier for further process

library(data.table) 

train_datatable = as.data.table(merged_sales)
monthly_shop_data = train_datatable[, list(item_cnt_month=(sum(item_cnt_day))/12), by = c("date_block_num", "month","shop_id", "item_category_id", "item_id", "item_price")]
summary(monthly_shop_data)
head(monthly_shop_data)
nrow(monthly_shop_data)

Attaching package: 'data.table'


The following objects are masked from 'package:lubridate':

    hour, isoweek, mday, minute, month, quarter, second, wday, week,
    yday, year


The following objects are masked from 'package:dplyr':

    between, first, last


The following object is masked from 'package:purrr':

    transpose

 date_block_num      month           shop_id        item_category_id
 Min.   : 0.00   1      :175576   31     : 109408   40     :343281  
 1st Qu.: 6.00   3      :167065   25     :  95955   55     :249575  
 Median :14.00   2      :163914   54     :  75612   30     :152805  
 Mean   :14.75   8      :152691   28     :  70733   37     :143685  
 3rd Qu.:23.00   6      :145226   57     :  64235   19     :113445  
 Max.   :33.00   7      :144487   42     :  62337   23     : 86916  
                 (Other):790063   (Other):1260742   (Other):649315  
    item_id          item_price       item_cnt_month     
 17717  :   7475   Min.   :    -1.0   Min.   : -1.83333  
 7893   :   2009   1st Qu.:   199.0   1st Qu.:  0.08333  
 2445   :   1993   Median :   399.0   Median :  0.08333  
 4244   :   1941   Mean   :   845.8   Mean   :  0.17482  
 6675   :   1856   3rd Qu.:   899.0   3rd Qu.:  0.16667  
 4248   :   1833   Max.   :307980.0   Max.   :180.75000  
 (Other):1721915

memory.size()
memory.limit()
memory.limit(size=56000)

We will divide our data into training(70%) and test data(30%)¶

library(caret)
set.seed(100)

# Divinding the data in 70/30 ratio into training and testing data
summarized_data = createDataPartition(y = monthly_shop_data$item_cnt_month, p = 0.7, list = F)
sum_training = monthly_shop_data[summarized_data, ]
testing = monthly_shop_data[-summarized_data, ]

Loading required package: lattice


Attaching package: 'caret'


The following object is masked from 'package:purrr':

    lift

We further divide the training data into training(80%) and validation(20%) data¶

This validation data helps us determine weather the trained model is good or not.

set.seed(100)

# Divinding the training data further in 80/20 ratio into training and validation data
training_data = createDataPartition(y = sum_training$item_cnt_month, p = 0.8, list = F)
training = sum_training[training_data,]
validation = sum_training[-training_data,]

nrow(training)
nrow(validation)
nrow(testing)

After trying and evaluating different models and different fetures we would decide to choose 'date_block_num', 'month', 'shop_id' and 'item_category_id' as our features as per the observations.¶

Training¶

model = train(item_cnt_month~date_block_num+month+shop_id+item_category_id, data=training, method="glm", trControl=trainControl(method="cv",number=5))

We are using the glm method for training and for trainControl we are using 5-fold cross validation.

summary(model)

Call:
NULL

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-12.588   -0.087   -0.034    0.016  179.965  

Coefficients: (2 not defined because of singularities)
                     Estimate Std. Error t value Pr(>|t|)    
(Intercept)         1.240e-01  3.865e-01   0.321 0.748275    
date_block_num     -1.157e-03  6.477e-05 -17.870  < 2e-16 ***
month2             -4.791e-03  2.511e-03  -1.908 0.056424 .  
month3              2.880e-03  2.507e-03   1.149 0.250607    
month4             -4.369e-03  2.627e-03  -1.663 0.096294 .  
month5             -1.238e-02  2.633e-03  -4.702 2.57e-06 ***
month6             -7.247e-03  2.613e-03  -2.773 0.005553 ** 
month7             -1.084e-02  2.622e-03  -4.132 3.59e-05 ***
month8             -1.066e-02  2.593e-03  -4.111 3.93e-05 ***
month9              1.467e-02  2.698e-03   5.435 5.47e-08 ***
month10             1.470e-02  2.729e-03   5.384 7.27e-08 ***
month11             3.087e-02  2.897e-03  10.659  < 2e-16 ***
month12             4.367e-02  2.659e-03  16.426  < 2e-16 ***
shop_id1           -4.429e-02  1.664e-02  -2.662 0.007779 ** 
shop_id2           -9.009e-02  1.205e-02  -7.477 7.59e-14 ***
shop_id3           -9.635e-02  1.197e-02  -8.049 8.37e-16 ***
shop_id4           -7.553e-02  1.157e-02  -6.527 6.71e-11 ***
shop_id5           -7.858e-02  1.159e-02  -6.777 1.23e-11 ***
shop_id6           -3.759e-02  1.112e-02  -3.380 0.000725 ***
shop_id7           -5.579e-02  1.132e-02  -4.928 8.29e-07 ***
shop_id8           -8.616e-02  1.783e-02  -4.833 1.34e-06 ***
shop_id9            4.151e-01  2.014e-02  20.610  < 2e-16 ***
shop_id10          -1.076e-01  1.214e-02  -8.864  < 2e-16 ***
shop_id11          -9.053e-02  3.653e-02  -2.478 0.013212 *  
shop_id12           7.741e-03  1.176e-02   0.659 0.510194    
shop_id13          -7.330e-02  1.249e-02  -5.871 4.33e-09 ***
shop_id14          -6.273e-02  1.161e-02  -5.401 6.63e-08 ***
shop_id15          -4.282e-02  1.134e-02  -3.777 0.000159 ***
shop_id16          -6.473e-02  1.135e-02  -5.701 1.19e-08 ***
shop_id17          -7.867e-02  1.211e-02  -6.496 8.23e-11 ***
shop_id18          -4.786e-02  1.136e-02  -4.214 2.51e-05 ***
shop_id19          -5.227e-02  1.126e-02  -4.644 3.42e-06 ***
shop_id20           3.378e-01  2.693e-02  12.541  < 2e-16 ***
shop_id21          -5.995e-02  1.126e-02  -5.323 1.02e-07 ***
shop_id22          -6.554e-02  1.153e-02  -5.685 1.31e-08 ***
shop_id23          -6.546e-02  1.512e-02  -4.328 1.51e-05 ***
shop_id24          -4.813e-02  1.140e-02  -4.220 2.44e-05 ***
shop_id25           2.060e-02  1.090e-02   1.890 0.058731 .  
shop_id26          -5.195e-02  1.132e-02  -4.587 4.49e-06 ***
shop_id27           1.083e-02  1.109e-02   0.977 0.328772    
shop_id28           2.207e-02  1.099e-02   2.009 0.044569 *  
shop_id29          -4.919e-02  1.145e-02  -4.297 1.73e-05 ***
shop_id30          -5.416e-02  1.135e-02  -4.771 1.83e-06 ***
shop_id31           4.827e-02  1.087e-02   4.442 8.91e-06 ***
shop_id32          -8.848e-02  1.432e-02  -6.177 6.53e-10 ***
shop_id33          -1.021e-01  1.615e-02  -6.320 2.61e-10 ***
shop_id34          -1.275e-01  1.537e-02  -8.295  < 2e-16 ***
shop_id35          -8.335e-02  1.126e-02  -7.402 1.34e-13 ***
shop_id36          -1.676e-01  4.567e-02  -3.669 0.000243 ***
shop_id37          -6.952e-02  1.151e-02  -6.038 1.56e-09 ***
shop_id38          -6.764e-02  1.147e-02  -5.895 3.74e-09 ***
shop_id39          -8.867e-02  1.288e-02  -6.884 5.82e-12 ***
shop_id40          -8.852e-02  1.860e-02  -4.760 1.94e-06 ***
shop_id41          -7.089e-02  1.152e-02  -6.154 7.58e-10 ***
shop_id42          -3.895e-03  1.104e-02  -0.353 0.724140    
shop_id43          -4.475e-02  1.163e-02  -3.847 0.000119 ***
shop_id44          -7.637e-02  1.150e-02  -6.642 3.10e-11 ***
shop_id45          -7.786e-02  1.162e-02  -6.703 2.05e-11 ***
shop_id46          -4.219e-02  1.125e-02  -3.750 0.000177 ***
shop_id47          -5.077e-02  1.135e-02  -4.474 7.68e-06 ***
shop_id48          -6.718e-02  1.229e-02  -5.465 4.63e-08 ***
shop_id49          -1.021e-01  1.263e-02  -8.089 6.03e-16 ***
shop_id50          -4.213e-02  1.126e-02  -3.742 0.000182 ***
shop_id51          -6.771e-02  1.142e-02  -5.931 3.02e-09 ***
shop_id52          -7.226e-02  1.146e-02  -6.305 2.89e-10 ***
shop_id53          -5.786e-02  1.140e-02  -5.077 3.83e-07 ***
shop_id54           1.369e-02  1.095e-02   1.250 0.211249    
shop_id55           1.129e+01  1.143e-01  98.761  < 2e-16 ***
shop_id56          -5.664e-02  1.117e-02  -5.071 3.95e-07 ***
shop_id57          -1.055e-02  1.103e-02  -0.956 0.338989    
shop_id58          -4.030e-02  1.125e-02  -3.583 0.000340 ***
shop_id59          -7.366e-02  1.153e-02  -6.388 1.68e-10 ***
item_category_id1          NA         NA      NA       NA    
item_category_id2   5.886e-02  3.866e-01   0.152 0.878979    
item_category_id3   1.644e-01  3.866e-01   0.425 0.670735    
item_category_id4   6.264e-02  3.870e-01   0.162 0.871422    
item_category_id5   3.656e-02  3.867e-01   0.095 0.924672    
item_category_id6   9.355e-02  3.866e-01   0.242 0.808806    
item_category_id7   7.516e-02  3.868e-01   0.194 0.845910    
item_category_id8   1.558e+00  3.882e-01   4.014 5.96e-05 ***
item_category_id9   6.766e-01  3.869e-01   1.749 0.080312 .  
item_category_id10         NA         NA      NA       NA    
item_category_id11  3.860e-02  3.867e-01   0.100 0.920491    
item_category_id12  2.768e-01  3.868e-01   0.716 0.474127    
item_category_id13  1.831e-02  3.889e-01   0.047 0.962446    
item_category_id14  2.769e-02  3.868e-01   0.072 0.942943    
item_category_id15  5.366e-02  3.867e-01   0.139 0.889648    
item_category_id16  5.917e-02  3.869e-01   0.153 0.878468    
item_category_id17  5.909e-03  3.888e-01   0.015 0.987873    
item_category_id18  7.367e-02  4.574e-01   0.161 0.872027    
item_category_id19  1.187e-01  3.865e-01   0.307 0.758776    
item_category_id20  2.589e-01  3.866e-01   0.670 0.502971    
item_category_id21  5.238e-02  3.866e-01   0.136 0.892209    
item_category_id22  6.543e-02  3.866e-01   0.169 0.865606    
item_category_id23  9.445e-02  3.865e-01   0.244 0.806970    
item_category_id24  7.239e-02  3.866e-01   0.187 0.851462    
item_category_id25  4.338e-02  3.866e-01   0.112 0.910653    
item_category_id26 -1.094e+01  4.286e-01 -25.518  < 2e-16 ***
item_category_id27 -1.130e+01  5.118e-01 -22.079  < 2e-16 ***
item_category_id28  1.921e-01  3.866e-01   0.497 0.619237    
item_category_id29  2.149e-01  3.866e-01   0.556 0.578271    
item_category_id30  1.809e-01  3.865e-01   0.468 0.639745    
item_category_id31 -1.109e+01  4.030e-01 -27.511  < 2e-16 ***
item_category_id32  2.469e-01  3.869e-01   0.638 0.523282    
item_category_id33  1.044e-01  3.867e-01   0.270 0.787188    
item_category_id34 -6.346e+00  4.100e-01 -15.479  < 2e-16 ***
item_category_id35  3.799e-01  3.867e-01   0.983 0.325835    
item_category_id36 -1.118e+01  4.713e-01 -23.727  < 2e-16 ***
item_category_id37  3.952e-02  3.865e-01   0.102 0.918569    
item_category_id38  3.783e-02  3.866e-01   0.098 0.922047    
item_category_id39  2.575e-02  3.879e-01   0.066 0.947075    
item_category_id40  7.569e-02  3.865e-01   0.196 0.844749    
item_category_id41  2.893e-02  3.866e-01   0.075 0.940336    
item_category_id42  2.162e-01  3.871e-01   0.559 0.576453    
item_category_id43  1.395e-03  3.866e-01   0.004 0.997121    
item_category_id44 -1.130e+01  4.059e-01 -27.836  < 2e-16 ***
item_category_id45 -2.143e-03  3.867e-01  -0.006 0.995579    
item_category_id46  3.667e-03  4.322e-01   0.008 0.993230    
item_category_id47  1.122e-01  3.868e-01   0.290 0.771780    
item_category_id48 -6.793e-03  4.990e-01  -0.014 0.989138    
item_category_id49  9.056e-02  3.866e-01   0.234 0.814795    
item_category_id50  1.358e-01  5.467e-01   0.248 0.803823    
item_category_id51  5.859e-02  6.695e-01   0.088 0.930255    
item_category_id52  5.183e-02  5.466e-01   0.095 0.924459    
item_category_id53 -3.406e-02  6.695e-01  -0.051 0.959418    
item_category_id54 -1.074e+01  4.037e-01 -26.603  < 2e-16 ***
item_category_id55  4.188e-02  3.865e-01   0.108 0.913716    
item_category_id56  1.612e-02  3.866e-01   0.042 0.966738    
item_category_id57  1.569e-02  3.866e-01   0.041 0.967621    
item_category_id58 -3.676e-04  3.866e-01  -0.001 0.999241    
item_category_id59 -1.082e-02  3.867e-01  -0.028 0.977678    
item_category_id60 -3.018e-03  3.870e-01  -0.008 0.993778    
item_category_id61  3.000e-02  3.866e-01   0.078 0.938149    
item_category_id62  5.092e-02  3.866e-01   0.132 0.895212    
item_category_id63  7.000e-02  3.866e-01   0.181 0.856299    
item_category_id64  5.896e-02  3.866e-01   0.153 0.878779    
item_category_id65  1.637e-01  3.866e-01   0.423 0.672006    
item_category_id66  1.042e-01  3.868e-01   0.269 0.787730    
item_category_id67  3.908e-02  3.866e-01   0.101 0.919472    
item_category_id68 -6.079e-02  3.930e-01  -0.155 0.877084    
item_category_id69  8.686e-02  3.866e-01   0.225 0.822222    
item_category_id70  1.290e-01  3.866e-01   0.334 0.738614    
item_category_id71  9.682e+00  3.870e-01  25.018  < 2e-16 ***
item_category_id72  5.562e-02  3.866e-01   0.144 0.885584    
item_category_id73  1.155e-01  3.867e-01   0.299 0.765192    
item_category_id74 -1.115e+01  4.305e-01 -25.889  < 2e-16 ***
item_category_id75  1.167e-01  3.866e-01   0.302 0.762799    
item_category_id76 -1.110e+01  4.034e-01 -27.518  < 2e-16 ***
item_category_id77  6.641e-02  3.868e-01   0.172 0.863696    
item_category_id78 -1.127e+01  4.034e-01 -27.935  < 2e-16 ***
item_category_id79  1.145e-01  3.866e-01   0.296 0.767151    
item_category_id80  1.207e+00  3.883e-01   3.108 0.001884 ** 
item_category_id81  1.662e-01  3.884e-01   0.428 0.668697    
item_category_id82  1.015e+00  3.873e-01   2.620 0.008790 ** 
item_category_id83  3.187e-01  3.868e-01   0.824 0.410037    
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

(Dispersion parameter for gaussian family taken to be 0.2987583)

    Null deviance: 382574  on 973854  degrees of freedom
Residual deviance: 290902  on 973702  degrees of freedom
AIC: 1587302

Number of Fisher Scoring iterations: 2

Evaluation¶

# Predicting the data from validation set
p = predict(model, validation)
error = (p - validation$item_cnt_month)

# Comparing the RMSE error from the model and the validation set

RMSE_model = model$results$RMSE; 
cat("RMSE error of the model:",RMSE_model)

RMSE_new = sqrt(mean(error^2)); 
cat("\nRMSE error of the predicted values:",RMSE_new)

RMSE error of the model: 0.5454389
RMSE error of the predicted values: 0.5224954

# Predicting the test set
prediction = predict(model, testing)

# Determine the error in the test set
test_error = (prediction - testing$item_cnt_month)
RMSE_test = sqrt(mean(test_error^2))
cat("RMSE error for the test data:",RMSE_test)

RMSE error for the test data: 0.5833928

These are accesptable values to pass this model. Obviously, we can improve on our model to reduce the error further but that will make out model more complex which is undesiarble.¶

Prediction¶

Now we imprt the test data from the CSV file from which we have the predict the data.

# import test data to be predicted
test_data = read_csv("test.csv")

# Assign 11 to month for November
test_data$month = 11
test_data$month = as.factor(test_data$month)
test_data$shop_id = as.factor(test_data$shop_id)

# Assign 34 to date_block_num for Nov 2015
test_data$date_block_num = 34
head(test_data)

Parsed with column specification:
cols(
  ID = col_double(),
  shop_id = col_double(),
  item_id = col_double()
)

#combine test_data and items_data by item_id to get item_category_id

combined_test = left_join(x=test_data, y = items, by = c("item_id"))
combined_test$item_category_id = as.factor(combined_test$item_category_id)
head(combined_test)

# Fit the model for our predictions
test_prediction = predict(model, combined_test)

# Prepart and write the predicted data as per the required format into a CSV file.
submit = cbind(combined_test[,1], test_prediction)
colnames(submit) = c("ID", "item_cnt_month")
class(submit)
submit$ID = as.integer(submit$ID)
head(submit)

write.csv(submit, "submission.csv", row.names = FALSE, quote = FALSE)

date	date_block_num	shop_id	item_id	item_price	item_cnt_day
<chr>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>
02.01.2013	0	59	22154	999.00	1
03.01.2013	0	25	2552	899.00	1
05.01.2013	0	25	2552	899.00	-1
06.01.2013	0	25	2554	1709.05	1
15.01.2013	0	25	2555	1099.00	1
10.01.2013	0	25	2564	349.00	1

shop_name	shop_id
<chr>	<dbl>
!<U+042F><U+043A><U+0443><U+0442><U+0441><U+043A> <U+041E><U+0440><U+0434><U+0436><U+043E><U+043D><U+0438><U+043A><U+0438><U+0434><U+0437><U+0435>, 56 <U+0444><U+0440><U+0430><U+043D>	0
!<U+042F><U+043A><U+0443><U+0442><U+0441><U+043A> <U+0422><U+0426> "<U+0426><U+0435><U+043D><U+0442><U+0440><U+0430><U+043B><U+044C><U+043D><U+044B><U+0439>" <U+0444><U+0440><U+0430><U+043D>	1
<U+0410><U+0434><U+044B><U+0433><U+0435><U+044F> <U+0422><U+0426> "<U+041C><U+0435><U+0433><U+0430>"	2
<U+0411><U+0430><U+043B><U+0430><U+0448><U+0438><U+0445><U+0430> <U+0422><U+0420><U+041A> "<U+041E><U+043A><U+0442><U+044F><U+0431><U+0440><U+044C>-<U+041A><U+0438><U+043D><U+043E><U+043C><U+0438><U+0440>"	3
<U+0412><U+043E><U+043B><U+0436><U+0441><U+043A><U+0438><U+0439> <U+0422><U+0426> "<U+0412><U+043E><U+043B><U+0433><U+0430> <U+041C><U+043E><U+043B><U+043B>"	4
<U+0412><U+043E><U+043B><U+043E><U+0433><U+0434><U+0430> <U+0422><U+0420><U+0426> "<U+041C><U+0430><U+0440><U+043C><U+0435><U+043B><U+0430><U+0434>"	5

item_category_name	item_category_id
<chr>	<dbl>
PC - <U+0413><U+0430><U+0440><U+043D><U+0438><U+0442><U+0443><U+0440><U+044B>/<U+041D><U+0430><U+0443><U+0448><U+043D><U+0438><U+043A><U+0438>	0
<U+0410><U+043A><U+0441><U+0435><U+0441><U+0441><U+0443><U+0430><U+0440><U+044B> - PS2	1
<U+0410><U+043A><U+0441><U+0435><U+0441><U+0441><U+0443><U+0430><U+0440><U+044B> - PS3	2
<U+0410><U+043A><U+0441><U+0435><U+0441><U+0441><U+0443><U+0430><U+0440><U+044B> - PS4	3
<U+0410><U+043A><U+0441><U+0435><U+0441><U+0441><U+0443><U+0430><U+0440><U+044B> - PSP	4
<U+0410><U+043A><U+0441><U+0435><U+0441><U+0441><U+0443><U+0430><U+0440><U+044B> - PSVita	5

date	date_block_num	shop_id	item_id	item_price	item_cnt_day	item_name	item_category_id	shop_name	item_category_name
<chr>	<dbl>	<dbl>	<dbl>	<dbl>	<dbl>	<chr>	<dbl>	<chr>	<chr>
02.01.2013	0	59	22154	999.00	1	<U+042F><U+0412><U+041B><U+0415><U+041D><U+0418><U+0415> 2012 (BD)	37	<U+042F><U+0440><U+043E><U+0441><U+043B><U+0430><U+0432><U+043B><U+044C> <U+0422><U+0426> "<U+0410><U+043B><U+044C><U+0442><U+0430><U+0438><U+0440>"	<U+041A><U+0438><U+043D><U+043E> - Blu-Ray
03.01.2013	0	25	2552	899.00	1	DEEP PURPLE The House Of Blue Light LP	58	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0420><U+041A> "<U+0410><U+0442><U+0440><U+0438><U+0443><U+043C>"	<U+041C><U+0443><U+0437><U+044B><U+043A><U+0430> - <U+0412><U+0438><U+043D><U+0438><U+043B>
05.01.2013	0	25	2552	899.00	-1	DEEP PURPLE The House Of Blue Light LP	58	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0420><U+041A> "<U+0410><U+0442><U+0440><U+0438><U+0443><U+043C>"	<U+041C><U+0443><U+0437><U+044B><U+043A><U+0430> - <U+0412><U+0438><U+043D><U+0438><U+043B>
06.01.2013	0	25	2554	1709.05	1	DEEP PURPLE Who Do You Think We Are LP	58	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0420><U+041A> "<U+0410><U+0442><U+0440><U+0438><U+0443><U+043C>"	<U+041C><U+0443><U+0437><U+044B><U+043A><U+0430> - <U+0412><U+0438><U+043D><U+0438><U+043B>
15.01.2013	0	25	2555	1099.00	1	DEEP PURPLE 30 Very Best Of 2CD (<U+0424><U+0438><U+0440><U+043C>.)	56	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0420><U+041A> "<U+0410><U+0442><U+0440><U+0438><U+0443><U+043C>"	<U+041C><U+0443><U+0437><U+044B><U+043A><U+0430> - CD <U+0444><U+0438><U+0440><U+043C><U+0435><U+043D><U+043D><U+043E><U+0433><U+043E> <U+043F><U+0440><U+043E><U+0438><U+0437><U+0432><U+043E><U+0434><U+0441><U+0442><U+0432><U+0430>
10.01.2013	0	25	2564	349.00	1	DEEP PURPLE Perihelion: Live In Concert DVD (<U+041A><U+0438><U+0440>.)	59	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0420><U+041A> "<U+0410><U+0442><U+0440><U+0438><U+0443><U+043C>"	<U+041C><U+0443><U+0437><U+044B><U+043A><U+0430> - <U+041C><U+0443><U+0437><U+044B><U+043A><U+0430><U+043B><U+044C><U+043D><U+043E><U+0435> <U+0432><U+0438><U+0434><U+0435><U+043E>

month	item_cnt_day
<fct>	<dbl>
1	359349
2	321975
3	344453
4	281573
5	277055
6	286924
7	271433
8	294091
9	305383
10	306019
11	247854
12	352097

shop_id	shop_name	number
<fct>	<chr>	<int>
0	!<U+042F><U+043A><U+0443><U+0442><U+0441><U+043A> <U+041E><U+0440><U+0434><U+0436><U+043E><U+043D><U+0438><U+043A><U+0438><U+0434><U+0437><U+0435>, 56 <U+0444><U+0440><U+0430><U+043D>	9857
1	!<U+042F><U+043A><U+0443><U+0442><U+0441><U+043A> <U+0422><U+0426> "<U+0426><U+0435><U+043D><U+0442><U+0440><U+0430><U+043B><U+044C><U+043D><U+044B><U+0439>" <U+0444><U+0440><U+0430><U+043D>	5678
2	<U+0410><U+0434><U+044B><U+0433><U+0435><U+044F> <U+0422><U+0426> "<U+041C><U+0435><U+0433><U+0430>"	25991
3	<U+0411><U+0430><U+043B><U+0430><U+0448><U+0438><U+0445><U+0430> <U+0422><U+0420><U+041A> "<U+041E><U+043A><U+0442><U+044F><U+0431><U+0440><U+044C>-<U+041A><U+0438><U+043D><U+043E><U+043C><U+0438><U+0440>"	25532
4	<U+0412><U+043E><U+043B><U+0436><U+0441><U+043A><U+0438><U+0439> <U+0422><U+0426> "<U+0412><U+043E><U+043B><U+0433><U+0430> <U+041C><U+043E><U+043B><U+043B>"	38242
5	<U+0412><U+043E><U+043B><U+043E><U+0433><U+0434><U+0430> <U+0422><U+0420><U+0426> "<U+041C><U+0430><U+0440><U+043C><U+0435><U+043B><U+0430><U+0434>"	38179
6	<U+0412><U+043E><U+0440><U+043E><U+043D><U+0435><U+0436> (<U+041F><U+043B><U+0435><U+0445><U+0430><U+043D><U+043E><U+0432><U+0441><U+043A><U+0430><U+044F>, 13)	82663
7	<U+0412><U+043E><U+0440><U+043E><U+043D><U+0435><U+0436> <U+0422><U+0420><U+0426> "<U+041C><U+0430><U+043A><U+0441><U+0438><U+043C><U+0438><U+0440>"	58076
8	<U+0412><U+043E><U+0440><U+043E><U+043D><U+0435><U+0436> <U+0422><U+0420><U+0426> <U+0421><U+0438><U+0442><U+0438>-<U+041F><U+0430><U+0440><U+043A> "<U+0413><U+0440><U+0430><U+0434>"	3412
9	<U+0412><U+044B><U+0435><U+0437><U+0434><U+043D><U+0430><U+044F> <U+0422><U+043E><U+0440><U+0433><U+043E><U+0432><U+043B><U+044F>	3751
10	<U+0416><U+0443><U+043A><U+043E><U+0432><U+0441><U+043A><U+0438><U+0439> <U+0443><U+043B>. <U+0427><U+043A><U+0430><U+043B><U+043E><U+0432><U+0430> 39<U+043C>?	21397
11	<U+0416><U+0443><U+043A><U+043E><U+0432><U+0441><U+043A><U+0438><U+0439> <U+0443><U+043B>. <U+0427><U+043A><U+0430><U+043B><U+043E><U+0432><U+0430> 39<U+043C>²	499
12	<U+0418><U+043D><U+0442><U+0435><U+0440><U+043D><U+0435><U+0442>-<U+043C><U+0430><U+0433><U+0430><U+0437><U+0438><U+043D> <U+0427><U+0421>	34694
13	<U+041A><U+0430><U+0437><U+0430><U+043D><U+044C> <U+0422><U+0426> "<U+0411><U+0435><U+0445><U+0435><U+0442><U+043B><U+0435>"	17824
14	<U+041A><U+0430><U+0437><U+0430><U+043D><U+044C> <U+0422><U+0426> "<U+041F><U+0430><U+0440><U+043A><U+0425><U+0430><U+0443><U+0441>" II	36979
15	<U+041A><U+0430><U+043B><U+0443><U+0433><U+0430> <U+0422><U+0420><U+0426> "XXI <U+0432><U+0435><U+043A>"	59511
16	<U+041A><U+043E><U+043B><U+043E><U+043C><U+043D><U+0430> <U+0422><U+0426> "<U+0420><U+0438><U+043E>"	52734
17	<U+041A><U+0440><U+0430><U+0441><U+043D><U+043E><U+044F><U+0440><U+0441><U+043A> <U+0422><U+0426> "<U+0412><U+0437><U+043B><U+0435><U+0442><U+043A><U+0430> <U+041F><U+043B><U+0430><U+0437><U+0430>"	22950
18	<U+041A><U+0440><U+0430><U+0441><U+043D><U+043E><U+044F><U+0440><U+0441><U+043A> <U+0422><U+0426> "<U+0418><U+044E><U+043D><U+044C>"	53227
19	<U+041A><U+0443><U+0440><U+0441><U+043A> <U+0422><U+0426> "<U+041F><U+0443><U+0448><U+043A><U+0438><U+043D><U+0441><U+043A><U+0438><U+0439>"	63911
20	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> "<U+0420><U+0430><U+0441><U+043F><U+0440><U+043E><U+0434><U+0430><U+0436><U+0430>"	1792
21	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+041C><U+0422><U+0420><U+0426> "<U+0410><U+0444><U+0438> <U+041C><U+043E><U+043B><U+043B>"	58133
22	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+041C><U+0430><U+0433><U+0430><U+0437><U+0438><U+043D> <U+0421>21	45434
23	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+041A> "<U+0411><U+0443><U+0434><U+0435><U+043D><U+043E><U+0432><U+0441><U+043A><U+0438><U+0439>" (<U+043F><U+0430><U+0432>.<U+0410>2)	6963
24	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+041A> "<U+0411><U+0443><U+0434><U+0435><U+043D><U+043E><U+0432><U+0441><U+043A><U+0438><U+0439>" (<U+043F><U+0430><U+0432>.<U+041A>7)	53032
25	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0420><U+041A> "<U+0410><U+0442><U+0440><U+0438><U+0443><U+043C>"	186104
26	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0426> "<U+0410><U+0440><U+0435><U+0430><U+043B>" (<U+0411><U+0435><U+043B><U+044F><U+0435><U+0432><U+043E>)	53910
27	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0426> "<U+041C><U+0415><U+0413><U+0410> <U+0411><U+0435><U+043B><U+0430><U+044F> <U+0414><U+0430><U+0447><U+0430> II"	105366
28	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0426> "<U+041C><U+0415><U+0413><U+0410> <U+0422><U+0435><U+043F><U+043B><U+044B><U+0439> <U+0421><U+0442><U+0430><U+043D>" II	142234
29	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0426> "<U+041D><U+043E><U+0432><U+044B><U+0439> <U+0432><U+0435><U+043A>" (<U+041D><U+043E><U+0432><U+043E><U+043A><U+043E><U+0441><U+0438><U+043D><U+043E>)	49225
30	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0426> "<U+041F><U+0435><U+0440><U+043B><U+043E><U+0432><U+0441><U+043A><U+0438><U+0439>"	50860
31	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0426> "<U+0421><U+0435><U+043C><U+0435><U+043D><U+043E><U+0432><U+0441><U+043A><U+0438><U+0439>"	235636
32	<U+041C><U+043E><U+0441><U+043A><U+0432><U+0430> <U+0422><U+0426> "<U+0421><U+0435><U+0440><U+0435><U+0431><U+0440><U+044F><U+043D><U+044B><U+0439> <U+0414><U+043E><U+043C>"	7947
33	<U+041C><U+044B><U+0442><U+0438><U+0449><U+0438> <U+0422><U+0420><U+041A> "XL-3"	5027
34	<U+041D>.<U+041D><U+043E><U+0432><U+0433><U+043E><U+0440><U+043E><U+0434> <U+0422><U+0420><U+0426> "<U+0420><U+0418><U+041E>"	5752
35	<U+041D>.<U+041D><U+043E><U+0432><U+0433><U+043E><U+0440><U+043E><U+0434> <U+0422><U+0420><U+0426> "<U+0424><U+0430><U+043D><U+0442><U+0430><U+0441><U+0442><U+0438><U+043A><U+0430>"	58445
36	<U+041D><U+043E><U+0432><U+043E><U+0441><U+0438><U+0431><U+0438><U+0440><U+0441><U+043A> <U+0422><U+0420><U+0426> "<U+0413><U+0430><U+043B><U+0435><U+0440><U+0435><U+044F> <U+041D><U+043E><U+0432><U+043E><U+0441><U+0438><U+0431><U+0438><U+0440><U+0441><U+043A>"	306
37	<U+041D><U+043E><U+0432><U+043E><U+0441><U+0438><U+0431><U+0438><U+0440><U+0441><U+043A> <U+0422><U+0426> "<U+041C><U+0435><U+0433><U+0430>"	39638
38	<U+041E><U+043C><U+0441><U+043A> <U+0422><U+0426> "<U+041C><U+0435><U+0433><U+0430>"	46013
39	<U+0420><U+043E><U+0441><U+0442><U+043E><U+0432><U+041D><U+0430><U+0414><U+043E><U+043D><U+0443> <U+0422><U+0420><U+041A> "<U+041C><U+0435><U+0433><U+0430><U+0446><U+0435><U+043D><U+0442><U+0440> <U+0413><U+043E><U+0440><U+0438><U+0437><U+043E><U+043D><U+0442>"	13440
40	<U+0420><U+043E><U+0441><U+0442><U+043E><U+0432><U+041D><U+0430><U+0414><U+043E><U+043D><U+0443> <U+0422><U+0420><U+041A> "<U+041C><U+0435><U+0433><U+0430><U+0446><U+0435><U+043D><U+0442><U+0440> <U+0413><U+043E><U+0440><U+0438><U+0437><U+043E><U+043D><U+0442>" <U+041E><U+0441><U+0442><U+0440><U+043E><U+0432><U+043D><U+043E><U+0439>	4257
41	<U+0420><U+043E><U+0441><U+0442><U+043E><U+0432><U+041D><U+0430><U+0414><U+043E><U+043D><U+0443> <U+0422><U+0426> "<U+041C><U+0435><U+0433><U+0430>"	41967
42	<U+0421><U+041F><U+0431> <U+0422><U+041A> "<U+041D><U+0435><U+0432><U+0441><U+043A><U+0438><U+0439> <U+0426><U+0435><U+043D><U+0442><U+0440>"	109253
43	<U+0421><U+041F><U+0431> <U+0422><U+041A> "<U+0421><U+0435><U+043D><U+043D><U+0430><U+044F>"	39282
44	<U+0421><U+0430><U+043C><U+0430><U+0440><U+0430> <U+0422><U+0426> "<U+041C><U+0435><U+043B><U+043E><U+0434><U+0438><U+044F>"	39530
45	<U+0421><U+0430><U+043C><U+0430><U+0440><U+0430> <U+0422><U+0426> "<U+041F><U+0430><U+0440><U+043A><U+0425><U+0430><U+0443><U+0441>"	35891
46	<U+0421><U+0435><U+0440><U+0433><U+0438><U+0435><U+0432> <U+041F><U+043E><U+0441><U+0430><U+0434> <U+0422><U+0426> "7<U+042F>"	66321
47	<U+0421><U+0443><U+0440><U+0433><U+0443><U+0442> <U+0422><U+0420><U+0426> "<U+0421><U+0438><U+0442><U+0438> <U+041C><U+043E><U+043B><U+043B>"	56695
48	<U+0422><U+043E><U+043C><U+0441><U+043A> <U+0422><U+0420><U+0426> "<U+0418><U+0437><U+0443><U+043C><U+0440><U+0443><U+0434><U+043D><U+044B><U+0439> <U+0413><U+043E><U+0440><U+043E><U+0434>"	21612
49	<U+0422><U+044E><U+043C><U+0435><U+043D><U+044C> <U+0422><U+0420><U+0426> "<U+041A><U+0440><U+0438><U+0441><U+0442><U+0430><U+043B><U+043B>"	15849
50	<U+0422><U+044E><U+043C><U+0435><U+043D><U+044C> <U+0422><U+0426> "<U+0413><U+0443><U+0434><U+0432><U+0438><U+043D>"	65173
51	<U+0422><U+044E><U+043C><U+0435><U+043D><U+044C> <U+0422><U+0426> "<U+0417><U+0435><U+043B><U+0435><U+043D><U+044B><U+0439> <U+0411><U+0435><U+0440><U+0435><U+0433>"	44433
52	<U+0423><U+0444><U+0430> <U+0422><U+041A> "<U+0426><U+0435><U+043D><U+0442><U+0440><U+0430><U+043B><U+044C><U+043D><U+044B><U+0439>"	43502
53	<U+0423><U+0444><U+0430> <U+0422><U+0426> "<U+0421><U+0435><U+043C><U+044C><U+044F>" 2	52921
54	<U+0425><U+0438><U+043C><U+043A><U+0438> <U+0422><U+0426> "<U+041C><U+0435><U+0433><U+0430>"	143480
55	<U+0426><U+0438><U+0444><U+0440><U+043E><U+0432><U+043E><U+0439> <U+0441><U+043A><U+043B><U+0430><U+0434> 1<U+0421>-<U+041E><U+043D><U+043B><U+0430><U+0439><U+043D>	34769
56	<U+0427><U+0435><U+0445><U+043E><U+0432> <U+0422><U+0420><U+0426> "<U+041A><U+0430><U+0440><U+043D><U+0430><U+0432><U+0430><U+043B>"	69573
57	<U+042F><U+043A><U+0443><U+0442><U+0441><U+043A> <U+041E><U+0440><U+0434><U+0436><U+043E><U+043D><U+0438><U+043A><U+0438><U+0434><U+0437><U+0435>, 56	117428
58	<U+042F><U+043A><U+0443><U+0442><U+0441><U+043A> <U+0422><U+0426> "<U+0426><U+0435><U+043D><U+0442><U+0440><U+0430><U+043B><U+044C><U+043D><U+044B><U+0439>"	71441
59	<U+042F><U+0440><U+043E><U+0441><U+043B><U+0430><U+0432><U+043B><U+044C> <U+0422><U+0426> "<U+0410><U+043B><U+044C><U+0442><U+0430><U+0438><U+0440>"	42108

	ID	item_cnt_month
	<int>	<dbl>
1	0	0.15569690
2	1	0.07887593
3	2	0.15569690
4	3	0.13144073
5	4	0.29592050
6	5	0.13144073

A tibble: 6 × 3
item_name	item_id	item_category_id
<chr>	<dbl>	<dbl>
! <U+0412><U+041E> <U+0412><U+041B><U+0410><U+0421><U+0422><U+0418> <U+041D><U+0410><U+0412><U+0410><U+0416><U+0414><U+0415><U+041D><U+0418><U+042F> (<U+041F><U+041B><U+0410><U+0421><U+0422>.) D	0	40
!ABBYY FineReader 12 Professional Edition Full [PC, <U+0426><U+0438><U+0444><U+0440><U+043E><U+0432><U+0430><U+044F> <U+0432><U+0435><U+0440><U+0441><U+0438><U+044F>]	1	76
***<U+0412> <U+041B><U+0423><U+0427><U+0410><U+0425> <U+0421><U+041B><U+0410><U+0412><U+042B> (UNV) D	2	40
***<U+0413><U+041E><U+041B><U+0423><U+0411><U+0410><U+042F> <U+0412><U+041E><U+041B><U+041D><U+0410> (Univ) D	3	40
***<U+041A><U+041E><U+0420><U+041E><U+0411><U+041A><U+0410> (<U+0421><U+0422><U+0415><U+041A><U+041B><U+041E>) D	4	40
***<U+041D><U+041E><U+0412><U+042B><U+0415> <U+0410><U+041C><U+0415><U+0420><U+0418><U+041A><U+0410><U+041D><U+0421><U+041A><U+0418><U+0415> <U+0413><U+0420><U+0410><U+0424><U+0424><U+0418><U+0422><U+0418> (UNI) D	5	40

A grouped_df: 6 × 3
item_id	item_name	freq
<fct>	<chr>	<int>
20949	<U+0424><U+0438><U+0440><U+043C><U+0435><U+043D><U+043D><U+044B><U+0439> <U+043F><U+0430><U+043A><U+0435><U+0442> <U+043C><U+0430><U+0439><U+043A><U+0430> 1<U+0421> <U+0418><U+043D><U+0442><U+0435><U+0440><U+0435><U+0441> <U+0431><U+0435><U+043B><U+044B><U+0439> (34*42) 45 <U+043C><U+043A><U+043C>	31340
5822	Playstation Store <U+043F><U+043E><U+043F><U+043E><U+043B><U+043D><U+0435><U+043D><U+0438><U+0435> <U+0431><U+0443><U+043C><U+0430><U+0436><U+043D><U+0438><U+043A><U+0430>: <U+041A><U+0430><U+0440><U+0442><U+0430> <U+043E><U+043F><U+043B><U+0430><U+0442><U+044B> 1000 <U+0440><U+0443><U+0431>.	9408
17717	<U+041F><U+0440><U+0438><U+0435><U+043C> <U+0434><U+0435><U+043D><U+0435><U+0436><U+043D><U+044B><U+0445> <U+0441><U+0440><U+0435><U+0434><U+0441><U+0442><U+0432> <U+0434><U+043B><U+044F> 1<U+0421>-<U+041E><U+043D><U+043B><U+0430><U+0439><U+043D>	9067
2808	Diablo III [PC, Jewel, <U+0440><U+0443><U+0441><U+0441><U+043A><U+0430><U+044F> <U+0432><U+0435><U+0440><U+0441><U+0438><U+044F>]	7479
4181	Kaspersky Internet Security Multi-Device Russian Edition. 2-Device 1 year Renewal Box	6853
7856	World of Warcraft. <U+041A><U+0430><U+0440><U+0442><U+0430> <U+043E><U+043F><U+043B><U+0430><U+0442><U+044B> <U+0438><U+0433><U+0440><U+043E><U+0432><U+043E><U+0433><U+043E> <U+0432><U+0440><U+0435><U+043C><U+0435><U+043D><U+0438> (online) (<U+0440><U+0443><U+0441>.<U+0432>.) (60 <U+0434><U+043D><U+0435><U+0439>) (Jewel)	6602

A grouped_df: 60 × 3
shop_id	item_id	quantity_sold
<fct>	<fct>	<dbl>
0	19811	79
1	13354	62
2	20949	1319
3	20949	1137
4	20949	2465
5	20949	2163
6	20949	3551
7	20949	3321
8	12168	31
9	7096	305
10	20949	1413
11	20949	46
12	11373	6563
13	20949	1023
14	20949	3219
15	20949	2907
16	20949	3395
17	20949	1377
18	20949	2826
19	20949	2347
20	15275	141
21	20949	4335
22	20949	4382
23	4164	89
24	20949	2918
25	20949	14343
26	20949	3624
27	20949	8265
28	20949	13458
29	20949	2167
30	20949	3922
31	20949	19934
32	2808	178
33	20949	379
34	20949	328
35	20949	3434
36	20949	16
37	20949	2727
38	20949	2096

A grouped_df: 6 × 3
item_category_id	item_category_name	freq
<fct>	<chr>	<int>
40	<U+041A><U+0438><U+043D><U+043E> - DVD	564652
30	<U+0418><U+0433><U+0440><U+044B> PC - <U+0421><U+0442><U+0430><U+043D><U+0434><U+0430><U+0440><U+0442><U+043D><U+044B><U+0435> <U+0438><U+0437><U+0434><U+0430><U+043D><U+0438><U+044F>	351591
55	<U+041C><U+0443><U+0437><U+044B><U+043A><U+0430> - CD <U+043B><U+043E><U+043A><U+0430><U+043B><U+044C><U+043D><U+043E><U+0433><U+043E> <U+043F><U+0440><U+043E><U+0438><U+0437><U+0432><U+043E><U+0434><U+0441><U+0442><U+0432><U+0430>	339585
19	<U+0418><U+0433><U+0440><U+044B> - PS3	208219
37	<U+041A><U+0438><U+043D><U+043E> - Blu-Ray	192674
23	<U+0418><U+0433><U+0440><U+044B> - XBOX 360	146789